Hemmings v. Padmore, 2026 SCC 34
Overview
In Hemmings, the Supreme Court of Canada (SCC) restated the separate roles of factual causation and remoteness in negligence. All three sets of reasons agree that the two inquiries are analytically distinct and that “remoteness” should replace “legal causation,” “proximate cause” and similar terms (paras. 43, 172, 289). The court divided, however, on how those principles applied to the trial judge’s reasons. The result left Dr. Jamensky, the anesthesiologist, solely liable; the actions against the prenatal obstetrician (Dr. Padmore) and the hospital (vicariously, for Nurse San Juan) remain dismissed. However, there is an unusual mathematical dilemma in the centre of the Padmore aspect of the case which is concerning. Adding up the votes, the plaintiff won on the issue of factual causation on a 6-to-1 basis. On the issue of remoteness, Chief Justice Wagner agreed with Justices Martin, Moreau and Karakatsanis that Dr. Padmore’s negligence was not too remote from the resultant injuries, resulting in a 4-3 decision in favour of the plaintiff. However, on the disposition, the Chief Justice sided with the Justices Rowe, O’Bonsawin and Jamal who upheld the order of the Court of Appeal. Problematically, however, that order was based on the Court of Appeal’s finding that Padmore’s negligence was too remote to the injuries, a conclusion specifically rejected by Chief Justice Wagner.
Facts
Sophia Hemmings, a morbidly obese, higher-risk patient, first attended Dr. Padmore in August 2008 seeking birth control. He did not order a pregnancy test until her fourth visit when she was already 12 weeks pregnant. He did not advise her of the risks associated with her pregnancy (including difficulty during delivery, likely requiring a C-section, and a higher risk of anesthetic complications and cardiac arrest), did not record those risks or a management plan and never raised the option of terminating the pregnancy (paras. 7–8).
On April 8, 2009, Sophia called the Hospital’s Labour and Delivery Unit complaining of vomiting. Nurse San Juan, who took the call, did not review her antenatal records or advise her to seek care (para. 10), did not take a proper history or fill in the form. On April 18, Dr. Padmore sent her for induction because she was now past due. However, labour stalled (one of the known risks associated with obesity in pregnancy) and Dr. O’Brien proceeded to perform a C-section under a spinal anesthetic on April 20, 2009. However, at the start of the procedure, Sophia panicked, so Dr. Jamensky, the anesthesiologist, converted her to general anesthesia. He needed two attempts to intubate. Her blood pressure, pulse and end-tidal CO₂ fell, and she suffered a cardiac arrest and permanent anoxic brain injury. The baby was delivered safely (paras. 12–16). Damages were agreed upon just prior to trial, so the trial proceeded on liability only (para. 1).
Decisions Below
Trial (Dow J.) found Drs. Padmore and Jamensky liable, and the hospital vicariously liable (Nurse San Juan) and directly liable (failure to convey records), apportioning damages equally among the three. Dr. Peng was not negligent. Dr. O’Brien breached the standard of care but was found not to have caused the injuries, which were also found to be too remote, in relation to her actions (para. 17). Dr. Padmore’s failures to advise of the risks and of termination were held to be factual “but for” causes, and cardiac arrest was held reasonably foreseeable to him (para. 19). Dr. Jamensky’s intubation and conversion breaches were the “direct cause” of the arrest. The amniotic fluid embolism theory, advanced by the defendants, was rejected (paras. 21–22).
Court of Appeal allowed the appeals of Dr. Padmore and the hospital and dismissed the actions against them. The court found the trial judge had not explained his preference between conflicting experts on what would have followed an April 8 referral (para. 24). The court also found what it called a “patent inconsistency”: if cardiac arrest from an anesthetic complication was unforeseeable to Dr. O’Brien at delivery, it could not have been foreseeable to Dr. Padmore or Nurse San Juan who interacted with the patient at a much earlier stage. Because the O’Brien finding was unappealed, it declined to order a new trial (para. 25). It set aside the intubation breach but upheld liability on the conversion breach (paras. 26–27).
Issues
- Did the trial judge err in his factual causation and remoteness analyses for Nurse San Juan and Dr. Padmore?
- Were the trial reasons sufficient, and was the finding that Dr. Jamensky’s conversion breach caused the injury sustainable?
Plurality Reasons: Rowe and O’Bonsawin JJ.
Factual causation. The plaintiff must prove on a balance of probabilities that, but for the negligent act, the injury would not have occurred. The act must be a necessary – not sole – cause, and the test is applied in a robust, common-sense way without requiring scientific precision or a definitive expert opinion (paras. 45–47). A finding of causation necessarily excludes the counterfactual in which the injury occurs anyway (para. 48). In simple cases the counterfactual may be inferred (para. 49). An explicit counterfactual analysis may be needed where there are multiple defendants, significant elapsed time, competing evidence-based theories or equivocal or contradictory expert evidence, as is common in medico-legal causation (paras. 50–51).
Remoteness. After reviewing the doctrine from Re Polemis through The Wagon Mound (No. 2), Côté and Mustapha (paras. 53–68), the plurality set out the governing principles. (1) Foreseeability is assessed at the time of breach, from the defendant’s vantage point; professionals are measured against the reasonable practitioner in their position (para. 70). (2) There must be a “real risk,” greater than a mere possibility, that a reasonable person would not brush aside as far-fetched and would take precautions against (para. 71). (3) The type of harm must be foreseeable, but not the precise concatenation of events (para. 72). Remoteness is not a free-floating public policy exercise. It is anchored in corrective justice and guided by precedent (paras. 75–78), and it weighs both the gravity and the probability of harm (para. 82). Liability cannot be set categorically by reference to a profession or activity (para. 84). The plurality preferred the orthodox reasonable-foreseeability framing to a “scope of risk” framing, while noting the two do not conflict (para. 85).
Standard of review. Factual causation is a finding of fact reviewable for palpable and overriding error, but failing to undertake a required counterfactual analysis is an error of law (para. 87). Remoteness is a question of mixed fact and law (para. 88).
Application to Nurse San Juan. The trial judge found only that an April 8 attendance would have led to induction or urgent C-section. He made no finding on whether either course would have avoided the same complications, and he did not resolve the equivocal expert evidence (paras. 91–93). Her telephone advice “directed Ms. Hemmings down the path” to injury rather than causing it physically, so an explicit counterfactual was required: were the risks the same on April 8 as on April 20? (paras. 94–95). Describing causation as “not ‘far-fetched’” imported the remoteness threshold into factual causation (para. 96).
Remoteness for Dr. Padmore and Nurse San Juan. The injury was properly characterized as cardiac arrest and resulting brain injury from anesthetic complications. That is neither so narrow as to require foresight of the precise sequence nor as broad as “physical injury” (paras. 99–105, distinguishing Michaluk). The trial judge erred in law by not asking whether a real risk of that injury was foreseeable to each defendant, given when each interacted with Sophia Hemmings and what each knew (para. 97). Remoteness is individualized, but findings about co-defendants cannot be logically incompatible (para. 120). Dr. O’Brien knew at least as much as Dr. Padmore (morbid obesity) and more (labour, vitals, potassium levels) yet was found unable to foresee the injury (paras. 122–124). The court would not disturb that unchallenged finding, because appeals are decided as the parties frame them and Dr. O’Brien was not a party (paras. 113–114). The evidence also did not support foreseeability. Literature put the added risk for morbidly obese patients at about 1.5% over a 1.5% baseline, and the experts disagreed on whether the risk even warranted disclosure. That standard of care evidence was directly relevant to foreseeability (paras. 125–127). The plaintiffs therefore failed to prove a foreseeable real risk (para. 129).
Dr. Jamensky. Reasons must explain “what” and “why” so as to permit appellate review, read functionally, in context and with the presumption that the judge knows the law (paras. 136–139). The Court of Appeal properly consulted the record without “backfilling” (para. 147). The trial judge’s finding that the arrest would not have occurred without the general anesthetic stood regardless of the set-aside intubation breach (para. 151). Uncontroverted evidence showed that panicking patients can almost always be calmed. Combined with the accepted evidence of Sophia’s sister and the trial judge’s finding that Dr. Jamensky’s contrary account was unreliable, the counterfactual was implicitly rejected. Given the close temporal and causal link, no explicit counterfactual analysis was required (paras. 153–159).
Martin J. (Dissenting in Part)
Martin J. agreed on the factual causation principles and on Dr. Jamensky and accepted that remoteness is mixed fact and law and not a “residual policy lever” (paras. 165, 188, 238). She grounded remoteness in correlativity: it works with the duty of care so that wrong and injury are “two sides of the same injustice” (paras. 181–185). Injuries are not too remote where their possibility “formed part of what made the defendant’s conduct unreasonable in the first place” (para. 185). She offered a functionally equivalent, risk-based formulation that asks what forms of harm made the conduct negligent and said trial judges do not err by using either test (paras. 203–207).
On Dr. Padmore, she read the O’Brien remark as part of the factual causation discussion, not a remoteness finding (para. 213). Different outcomes for differently situated defendants are not legal error, because Dr. O’Brien’s breach (failure to assess) differed in kind from Dr. Padmore’s non-disclosure (paras. 214–215). The trial judge’s blurred language did not amount to reviewable error (paras. 217–224). The Court of Appeal framed the risk too narrowly by treating cardiac arrest at C-section as “extremely rare” (para. 229). The plaintiffs’ “physical injury” framing was too broad (para. 230). The proper risk was that elevated pregnancy-related risks, including anesthetic complications and cardiac arrest, would materialize. The injury fell squarely within it, a conclusion reinforced by Dr. Padmore’s actual knowledge of those risks and by analogy to informed consent under Reibl v. Hughes (paras. 231–236). She agreed the hospital claim failed: there was no discernible remoteness analysis, and the vomiting complaint bore no relationship to the later anesthetic event (paras. 239–245).
Wagner C.J. (Dissenting in Part)
The Chief Justice endorsed the plurality’s restatement of the law (para. 250) but held that the trial judge “fundamentally conflated” factual causation and remoteness throughout his 19 paragraphs on causation (paras. 253–258). The “far-fetched” language tainted the finding that Sophia Hemmings would have terminated the pregnancy, so the claim against Dr. Padmore failed on factual causation (paras. 260–263). He nevertheless agreed with Martin J. that the injuries were not too remote from Dr. Padmore. In his view, the O’Brien remark was itself a legal error, because remoteness “does not turn on whether the defendant could predict another tortfeasor’s negligence,” and an inconsistency must be resolved by testing each finding (paras. 264–270). On Dr. Jamensky, the trial judge applied “but for” to the conversion itself rather than to the breach (failing to exhaust alternatives) and never resolved the conflicting evidence on whether she would have calmed (paras. 277–283). He observed that the errors “were not the fault of plaintiff counsel” (para. 284).
Significance
- Vocabulary and structure. Plead, argue and write reasons treating factual causation (“but for,” balance of probabilities) and remoteness (real risk, reasonable foreseeability) as separate steps. “Far-fetched” belongs only to remoteness (paras. 43, 96).
- Counterfactuals for upstream defendants. Where a defendant’s breach (advice, triage, records) precedes the injury-causing event, lead evidence on and seek an express finding about what would have happened with proper care, including whether the same risks would have existed (paras. 50–51, 94–95).
- Characterizing the injury. “Physical injury” is too broad. Frame the type of harm at the level the evidence on risk can support (paras. 99–104). Standard of care evidence on whether a risk was worth disclosing may decide foreseeability (para. 127).
- Co-defendant findings. Unappealed findings about a co-defendant can bind the analysis for others (paras. 113–114, 124). Consider whether such findings should be challenged on appeal.
- Split on remoteness. Four judges (Martin J. with two others and Wagner C.J.) would have found the injuries not too remote from Dr. Padmore and rejected using the O’Brien finding to decide his case (paras. 228, 237, 264–269). The claim failed against Dr. Padmore only because Wagner C.J. found factual causation unproven. The risk-based approach in Martin J.’s reasons is a minority view, but it remains a persuasive tool for non-disclosure cases.
Thank you to OTLA members Jan Marin, Paul Harte and Richard Halpern for your work and intervention!
Congratulations to OTLA members Amani Oakley, Duncan Embury, Neil Oakley, Daniela Pacheco and Michelle Kudlats for their success in maintaining the verdict against Dr. Jamensky.